Thank you for your interest in contributing to the capsnet-nlp project. This guide details how to contribute in a way that is efficient for everyone.
Whenever you found a problem or want to file a bug report please use the capsnet-nlp issue tracker here.
Please search the issue tracker first before submitting your own. A bug report should be described as detailed as possible, e.g. if you write a bug report please provide detailed steps to reproduce the bug or provide relevant logs and error messages.
All issues should be written in English, but you're welcome to write in German.
We welcome merge requests with fixes and improvements to the capsnet-nlp code, tests, and/or further documentation.
If you can, please submit a merge request with the fix or improvements including unittests. In general bug fixes that include a regression test should be merged quickly, while new shiny features without proper unittests are very likely to to be merged later. The workflow to make a merge request is:
-
Fork the capsnet-nlp repository from GitHub.com:
https://github.com/stefan-it/capsnet-nlp.git -
Create a new feature branch, e.g. labeled with
feature/meaningful-description. If there's any issue available, please include the issue number in the branch name like:feature/issue-42 -
Write code/bug fixes new features and please also include test cases for them
-
Commit often and push to your feature branch
-
Provide a merge request on GitHub.com
-
In this merge request you should describe all implemented features and please link any relevant issues.
-
Be prepared to answer questions and incorporate feedback after your submission
Please keep the change in a single merge request as small as possible.
Please refer to our coding style document when writing code in Python.
This guide is heavily inspired by the excellent GitLab constributing guide, which is available here.